The present invention is related to a computer system and a method to assist an analysis of asynchronous remote replication.
Business can be affected greatly when an operation is suspended or a data loss occurs due to an occurrence of an IT system down. Disaster recovery (DR) as a preparation for unforeseeable events such as natural disasters is vital. A DR system is operable to duplicate data from a primary site to a secondary site at a remote location in an asynchronous manner. This is referred to as an asynchronous remote replication.
A synchronous remote replication notifies a host with a completion of a write after a completion of data replication to a storage apparatus (secondary storage apparatus) at a secondary site concerning a write access from the host to a storage apparatus (primary storage apparatus) at a primary site. When the completion of a write is notified, data at the primary site and that at the secondary site are matched.
On the other hand, the asynchronous remote replication notifies a host with a completion of a write when data writing to a primary storage apparatus is completed without waiting for the replication of the data at the secondary storage apparatus. In an environment where the asynchronous remote replication is used, a phenomenon where a delay occurs between the storing of data at the primary storage apparatus and the replicating of the data at the secondary storage apparatus causing the data at the secondary storage apparatus to be older than the data at the primary storage apparatus may occur.
An index for a maximum allowable time of data that is lost when a system at the primary site is down is referred to as RPO (Recovery Point Objective: Target Recovery Point). Further, an index for an amount of time needed for a system which is down to recover is referred to as RTO (Recovery Time Objective: Target Recovery Time).
These index values are linked to the survival of a business entity and are thus considered noteworthy as an SLA (Service Level Agreement). It is important that the asynchronous remote replication satisfies the demand with respect to the above mentioned index for the SLA.
The asynchronous remote replication appropriately transfers data from a primary storage apparatus to a secondary storage apparatus in an asynchronous manner by using, for example, an order guaranteeing technology of data. A delay time between a writing to the primary storage apparatus and that to the secondary storage apparatus changes depending on a load on a system which changes with time. For example, Patent Document 1 discloses a calculation method of a delay time concerning the asynchronous remote replication.
An example of a delay time will be described. A host writes data to a primary storage apparatus. The primary storage apparatus writes the data received from the host to a primary volume. The primary storage apparatus transfers the data from the host to a secondary storage apparatus. When transferring data, both the primary storage apparatus and the secondary storage apparatus temporarily store the data at a volume which is referred to as a journal volume.
For example, suppose that data, which is written to the primary storage apparatus at 8:00, is written to the secondary storage apparatus at 8:08. The delay time here is 8 minutes. Further, in this case, recoverable data as of 8:08 is the data written to the primary storage apparatus up until 8:00 (data recovery available time is 8:00). Accordingly, in the asynchronous remote replication, a delay occurs in writing data to the secondary storage apparatus with respect to data writing to the primary storage apparatus.
Patent Document 1: US 2006/0277384 A1